178 PRESSURE. 



In every month throughout the year there is more or less marked evidence of two 

 maxima and two minima. On the mean of the year (bottom but one curve) there i? a 

 minimum near 4 hours and a maximum at 10 hours, these will be referred to as the morning 

 minimum and maximum. The second minimum occurs at 16 hours and tl»e second maximum 

 at 20 hours, these will be called the evening minimum and maximum. 



During the months November, December and January the sun was above the horizon 

 the whole time. It will be noticed that the curves for January and December are very 

 similar, but that the curve for November is somewhat different. This difference is due to 

 1911 and 1912, for in 1902 and 1903 the November curve was almost exactly the same 

 type as that for January and December. It is very probable that with a .sufficient number 

 of years of observations the variation for November would prove to be similar to that of 

 January and December. The combined curve for these three months, shown as the summer 

 curve in the lower half of the figure, iii probably typical of the period during which the .sun 

 is above the horizon the whole time. In this curve the morning minimum and maximum 

 are highly develojied while the evoning minimum and maximum are almost absent. 



During February and October the sun only sets for a short period during a few of the 

 days, and these two months have remarkably similar daily pressure variations wh.ich are not 

 very different from those of March and September, when the days and nights are of approxi- 

 mately equal lengths. All the four curves, except for September, show the double daily varia- 

 tion in which the morning maximum is larger than the evening one. 



The sun rises only for a short time on some of the days in April and August, and is 

 below the horizon on the remaining days. These two months have almost identical variations, 

 whicli are different from those of the months on either side. In both months the morning 

 minimum and the evening maximum are abnormally developed, so that, except for a slight 

 irregularity at midday, the c^lrves are simple with a minimum at 6 hours and a maximum 

 at 20 hours, the amplitude being abnormally large. It is difficult to see any reason why the 

 appearance of the sun for a short time at midday should produce this effect on the pressure 

 variation. 



The sun is entirely below the horizon during May, June and July. It is therefore sur- 

 prising to fuid that while the curves for May and July are similar, the curve for June is 

 different from them. 



The following points appear of importance in this survey : — 



(a) The great similarity of the curves placed together in pairs in figure fjQ shows that 

 the daily barometer variation depends largely on the relative length of day and 

 night. 

 (h) During the months that the sun is more above the horizon than below, October 

 to February, the morning maximum is more developed than the evening one — 

 the chief feature of all the curves (the November curve qualified as stated above) 

 being the pronounced maximum at 10 hours, 

 (c) Three of the five months during which the sun is more below than above the 

 horizon, April, June and August, show an almost total absence of the morning 

 maximum ; but, on the other hand, the intermediate months. May and July, have 

 the morning maximum well developed and larger than the evening one. 

 It is difficult to combine {b) and (c) into a single statement giving the dependence of 

 the type of variation on the position of the sun. If it were not for May and July, it would 

 appear that the morning maximum was developed with a high sun and absent when the 

 sun was low, or entirely below the horizon ; but as the morning maximum is better developed 

 than the evening one during May and July, this generalisation does not hold. 



